Open Stack

Hi everybody,

First, some background:

A year and a half ago, Infra started down the road of of writing a
replacement for the pieces of Launchpad that OpenStack continues to use.
There were several reasons, but notable amongst them are:

- Desire to use the forthcoming openstackid OpenID/Oauth as an SSO
- Delay in long-standing bugs that affect OpenStack getting fixed

The existing open source offerings that we investigated did not have
adequate feature parity in the key data model areas that made Launchpad
particularly compelling as a choice for us, and adding what we needed to
the existing offerings would amount to substantial rewrites ... so we
decided that we had no real choice but to write our own.

Where we're at

We've gotten far enough to get Infra moved on to storyboard, but the
project has never really gotten resourced to the level it needs to be to
truly responsive to the needs of our community. We're making good
progress towards meeting the initial set of goals we set, but in the
mean time several new requests have come in - such as from the UX team
and the Product Management Working Group - that we cannot meet today and
which at our current rate I do not believe we would be able to meet in a
reasonable timeframe.

At the same time, the state of the art around us has improved since we
started. A year and a half ago, I was able to very honestly say that we
needed to work on this effort because we simply had no other choice.
That is no longer true. Existing Open Source offerings not only can
represent a large portion of our data needs, but additionally can
support the additional features that have been requested by our
community today out of the box.

The combination of the two of those makes the likelihood of us being
able to convince people to pony up more resources seem more and more far
fetched. I could be wrong, of course - it's possible that in response to
this someone will start jumping up and down and commit engineers to the
effort ... but I'm not holding my breath.

Biting the bullet

I think we should get out of the business of writing our own bug tracker.

It's not an easy thing to say, and I don't say it lightly. There are
things that storyboard models well that continue to be things that
simply are not modeled elsewhere. However, I think it's important to
know when good enough will do, and I think it's important to be able to
step up and say that we tried valiantly, and everyone involved did a
great job, and yet the world has moved on and writing a bug tracker is
not, at the end of the day, what we're all here to do.

We're looking at what our options are, and Thierry is examining them to
see how tolerable their differences would be to our community.

I propose that we have a solid answer and migration plan to put in front
of people by Vancouver at the latest.

Finally, I'd like to say thank you to the storyboard team for attacking
a very hard problem with not enough resources.

Monty